Kali Linux Bleeding Edge MCP Server w. SandBox & PowerShell

Kali Linux Bleeding Edge MCP Server w. SandBox & PowerShell

Runs Kali Linux security tools inside a sandbox and through PowerShell scripts, behind an MCP server and a Gradio page that HuggingChat can use. It suits security testers working on systems they own.

Made by DarkDriftz on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This Space is an MCP server and a Gradio app in one codebase, built around a collection of Kali Linux security tools. Its README claims more than 793 tools, 150 of them experimental. Version 4.0.0 adds a Linux sandbox that runs the tools in isolation and a PowerShell layer for scripted security checks.

Who it suits: It suits security testers, system administrators and students who want security tools reachable from an AI chat or a web page.

Industry | Who uses it | How they use it | Example
Security testing | Penetration testers | Run approved scans from an AI client or the web page | A tester asks the server to scan a lab machine for open ports.
IT operations | System administrators | Run audits and log checks with the PowerShell scripts | An admin runs the network audit script on a server they manage.
Compliance | Auditors | Check systems against CIS, NIST, GDPR and PCI-DSS with one script | An auditor runs the compliance check before a review.
Education | Security students | Practise with tools inside an isolated session | A student creates a session and runs one scan inside it.

If you have no permission to test the target systems, skip it. So can anyone who wants a small tool, since the repository holds 22 files.

Where to use it:

• A web browser, through the Space's page on Hugging Face
• HuggingChat and other MCP clients, over a live SSE connection
• A local copy on Windows, macOS or Linux, using Python

Use case | Who | What they get back
Scans inside a sandbox | Security testers | Tool output from a session with limited CPU, memory and processes
Scripted audits | Administrators | The results of five PowerShell security scripts
Service setup | Linux administrators | A generated systemd unit file for a PowerShell service

How to set it up:

• Python and pip on Windows, macOS or Linux
• The packages in requirements.txt, since the repository also has a smaller requirements-minimal.txt

1. Install the huggingface_hub package.
2. Download the Space into a local folder.
3. Move into that folder.
4. Install gradio at version 6.1.0.
5. Install the packages listed in requirements.txt.
6. Start the app.

pip install -U huggingface_hub
hf download DarkDriftz/Kali-Linux-Bleeding-Edge-MCP-Server --repo-type space --local-dir Kali-Linux-Bleeding-Edge-MCP-Server
cd Kali-Linux-Bleeding-Edge-MCP-Server
pip install gradio==6.1.0
pip install -r requirements.txt
python app.py

How to use it: Open the Space's page, or start the app locally with the last command above. Then call the tools from the page or from an MCP client such as HuggingChat. The Space's own README is the place to check the connection steps. In code, a sandbox session is created, used and destroyed like this.

# Create sandbox session
session_id = sandbox_executor.create_session()

# Execute command in sandbox
result = await sandbox_executor.execute_command(
    session_id=session_id,
    command="nmap -sV -p 1-1000 localhost",
    timeout=300
)

# Destroy session when done
sandbox_executor.destroy_session(session_id)

Tool | What it does
create_session | Opens a new isolated sandbox session
execute_command | Runs one approved tool inside a session, with a timeout
destroy_session | Ends the session and purges its folder
Invoke-VulnerabilityScan | Runs a vulnerability assessment in several phases
Invoke-PortScan | Scans ports with several threads at once
Invoke-NetworkAudit | Audits the network configuration
Invoke-LogAnalysis | Reads security event logs and looks for anomalies
Invoke-ComplianceCheck | Checks a system against CIS, NIST, GDPR and PCI-DSS

Example: You create a sandbox session and run an nmap version scan of ports 1 to 1000 on localhost, with a five minute limit. The command's scan output comes back to you. When you destroy the session, its folder is purged.

For newcomers and experts: If you are new, security tools are programs that probe a computer for weak spots. Here they run in a walled-off session. A safe first step is to open the Space's page and read the list of tools.

If you know the field, the sandbox uses user, PID and network namespaces and sets no_new_privileges. It caps memory at 512 MB, CPU at 50 percent and 20 processes. A whitelist of more than 30 tools limits what can run, and a session times out after 300 seconds by default.

A page or MCP client sends a request
The tool must be on the whitelist
A sandbox session runs it with limits
Output returns to the caller
The session folder is purged

Good to know:

• Run scans only against systems you own or may test
• The whitelist includes password cracking and web attack tools such as hydra, hashcat and sqlmap
• Many of the 22 repository files are Markdown reports and guides
• License: MIT
• Cost to run: none named
• Data it reaches: the systems and networks you point the tools at, plus files in sandbox session folders

On Hugging Face, 13 November 2026

• Likes: 2
• Built with: gradio
• License: mit

Links

• Space: https://huggingface.co/spaces/DarkDriftz/Kali-Linux-Bleeding-Edge-MCP-Server
• Files: https://huggingface.co/spaces/DarkDriftz/Kali-Linux-Bleeding-Edge-MCP-Server/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.